Commuters’ perception of safety in public transport: a review of approaches, gaps and emerging research opportunities

Commuter priorities in public transport have shifted from mere service availability and cost–benefit considerations to include perception of safety, which has become a key determinant of ridership and overall satisfaction. This review critically examines how commuters’ safety perceptions are studied in public transport and evaluates current methodological approaches, including the use and potential of novel data sources and machine learning (ML) techniques. A systematic review of 54 studies reveals three key research gaps. First, existing studies rely on conventional mapping and analysis methods, with limited use of novel data sources and ML techniques in geospatial analysis. This restricts the ability to capture non-linear relationships, uncover latent variables, and integrate multimodal data relevant to safety perception. Second, studies often overlook a comprehensive analysis of physical and environmental factors, despite the literature consistently identifying them as critical impacts on commuters’ safety perceptions, thus limiting the understanding of the role of the built environment. Third, reliance on traditional approaches for adopting theoretical frameworks restricts methodological innovation and limits the diversity of contexts addressed, reducing the scalability of existing studies. While ML techniques have demonstrated value in multiple urban studies, their potential remains underutilised in public transport safety research. Addressing these gaps through theory-driven, data-integrated approaches is vital to advancing a more inclusive and responsive public transport environment.